The detection of glycine from the treatment of glyoxylic acid with iron(II) sulfate and ammonia in water

Article abstract of DOI:10.3184/174751911X12964930076809

A glycine/(NH₄)₂SO₄ mixture was isolated by treatment of glyoxylic acid with FeSO₄ and aqNH₃ in H₂O. The yields of glycine were estimated by ¹H NMR. Pyruvic acid was not reduced to alanine under these conditions. This method for forming glycine might have occurred prebiotically alongside the Urey-Miller arc discharge method for making amino acids because glyoxylic acid is formed by arc discharge through a N₂/CO₂ atmosphere and both NH3 and Fe(II) occurred in the earth's early oceans. The carboxylic acid directs the reduction of 2,4-dinitrobenzoic acid to give 2-amino-4-nitrobenzoic acid.

Full text of DOI:10.3184/174751911X12964930076809

JOURNAL OF CHEMICAL RESEARCH 2011
RESEARCH PAPER 129
MARCH, 129–132
The detection of glycine from the treatment of glyoxylic acid with iron(II)
sulfate and ammonia in water
M. John Plater* and Ken Vassiliev
Department of Chemistry, University of Aberdeen, Meston Walk, Aberdeen AB24 3UE, UK
A glycine/(NH4)2SO4 mixture was isolated by treatment of glyoxylic acid with FeSO4 and aqNH3 in H2O. The yields of
glycine were estimated by 1H NMR. Pyruvic acid was not reduced to alanine under these conditions. This method for
forming glycine might have occurred prebiotically alongside the Urey-Miller arc discharge method for making amino
acids because glyoxylic acid is formed by arc discharge through a N2/CO2 atmosphere and both NH3 and Fe(II)
occurred in the earth’s early oceans. The carboxylic acid directs the reduction of 2,4-dinitrobenzoic acid to give
2-amino-4-nitrobenzoic acid.
Keywords: glyoxylic acid, iron(II) sulfate, ammonia, amino acid, ninhydrin, arc discharge
through carbon dioxide and nitrogen.5 Ammonia can be formed
by the reduction of nitrite with Fe(II).6–7 In persuit of further
potential prebiotic routes for making amino acids we exam-
ined the reductive amination of glyoxylic acid with Fe(II)SO4
7H2O and NH3 in H2O. α-Keto acids have previously been con-
verted to protonated α-imino acids and reduced with an NADH
analogue in low yields but not in water.10 Glyoxylic acid has
been reduced to glycine with Pd/H2 and aqNH3.11 α-Keto acids,
but not glyoxylic acid, have been reduced to amino acids using
precipitated FeS and Fe(OH)2 and amines.12 Fe(II)SO4 7H2O
was first used to reduce the nitro group with Ba(OH)2 in 188013
then with aqNH3 in 188214 and later in 191715 but has not
been used for reductive aminations. [Fe(II)SO4 7H2O has been
called green vitriol, copperas and melanterite. In the early
German papers, it is called ‘Eisenvitriol’.]
In 1938, an English translation of Oparin’s 1936 Russian book
‘Origin of Life’was published.1 One of Oparin’s key proposals
was that early earth had a reducing atmosphere of methane,
ammonia, hydrogen and water. He suggested methane and
ammonia as constituents because they are present on Jupiter
and Saturn and both hydrogen and water are ubiquitous.
In 1953 a gas mixture of this composition was subjected to an
arc discharge now known as the Urey–Miller experiment.2
This gave some amino acids which were detected by TLC after
spraying with ninhydrin. However, later research has sug-
gested that the earth’s early atmosphere was more likely to be
composed of carbon dioxide and nitrogen with a partial pres-
sure of carbon dioxide as high as 10 bars.3 This is based on
geological evidence since the absence of continents on early
earth meant that carbon was not stored as sedimentary carbon-
ate so more may have been in the atmosphere. The occurrence
of present day carbonates, chalk, limestone and dolomite,
as carbon dioxide was also proposed by Haldane in 1929.4 A
second arc discharge experiment was performed by Miller and
Lazcano using an atmosphere of carbon dioxide and nitrogen
to investigate a non reducing atmosphere.5 Provided a reducing
agent such as ascorbate or Fe(II) was present amino acids were
formed. Other key products were nitrite, nitrate, formaldehyde,
glycoaldehyde and glyoxylic acid. These compounds might
therefore have been present as a flux into the early earth’s
oceans. In 1993 Summers and Chang provided a possible
prebiotic route to ammonia by studying the reduction of nitrite
with Fe(II).6–7
Our studies were specifically to focus on water as solvent as
this could mimic a prebiotic solvent of early earth. Possible
roles of glyoxylic acid in prebiotic chemistry have been
reviewed.16
Discussion
Glyoxylic acid hydrate and Fe(II)SO4 7H2O were treated with
ammonia and left in water for 24 h at 75 °C and for 24 h at
room temperature (Scheme 1). Fe(II)SO4 7H2O is not a strong
reducing agent but complexation to ammonia makes it electron
rich so it readily donates an electron forming an Fe(III) salt.
We considered two possible reaction pathways. Firstly, the
reaction of glyoxylic acid and ammonia to form an imine.
We considered that this reversibly formed imine could com-
plex to Fe(II) followed by reduction. The imine is conjugated
to a carbonyl group which should facilitate electron transfer
from Fe(II).
Secondly, if Fe(II)(NH3)x(H2O)y [x + y = 6] could complex to
the glyoxylic acid carboxyl group and deliver ammonia to the
electrophilic aldehyde to give the Fe(II) glyoxylic acid imine
complex (shown in Scheme 1), the formation of an unbound,
hydrolytically unstable imine need not occur. This could be
helpful under dilute prebiotic conditions. Although the stoi-
chiometric reaction requires 2 equiv. of Fe(II) it was found that
a cleaner product filtrate was observed using only 1 equiv. The
reaction was worked up by carefully decanting firstly the still
solution onto silica followed by the iron salts, washing with
6 Fe(II) + 7H+ + NO2– = 6Fe(III) + 2H2O + NH3
The earth’s early oceans are considered to have been rich in
Fe(II) salts, a few mg L−1, which is 100 times more concen-
trated than today.8 The evidence for this comes from studies of
banded iron-formations (BIF) such as that of the Hamersley
Basin of Australia which has a circumference of 1000 miles.
The deposition of Fe(III), which is less soluble than Fe(II),
occurred at such a rate that weathering and transport of Fe(II)
by rivers to the ocean is unlikely. A number of BIF’s are
known.9 With the presence of Fe(II) and ammonia in the earth’s
early oceans it seems reasonable to study the reductive chem-
istry of the arc discharge products from a CO2/N2 atmosphere.
Both nitrite and glyoxylic acid can be formed by arc discharge
Scheme 1
* Correspondent. E-mail: m.j.plater@abdn.ac.uk

130 JOURNAL OF CHEMICAL RESEARCH 2011
Scheme 2
water and concentration in vacuo. The iron salts block the
filtration on silica. Both reactions gave similar analyses which
are now described. Analysis by TLC on a silica plate (2.5 cm
plate width, eluant 10% conc NH3/MeOH, development with a
ninhydrin dip of 1 mg mL−1 MeOH) showed a strong glycine
spot to be present compared with a glycine standard. After
running the silica plate in conc NH3/MeOH solution it was
dried at 100 °C for 1 min to remove traces of ammonia. If the
reaction itself is analysed with less than 10% NH3 in the eluent
(which might be owing to evaporation) the glycine streaks
but this is caused by the iron salts and was demonstrated by
analysing a mixture of glycine with some Fe(II)SO4 7H2O.
Glycine from the filtered solution did not streak on the TLC
plate. Analysis of the IR spectrum showed a strong carbonyl
stretch at 1600–1700 cm−1 the same as that for glycine. Glyox-
ylic acid showed a much higher carbonyl stretch between 1700
and 1760 cm−1 which proved that the carbonyl stretch was not
from glyoxylic acid. Three broad resonances at 1380–1410,
1080–1120 and 600–620 cm−1 match those of (NH4)2SO4
confirming that this is present in the sample. Analysis of the
sample by UV confirmed that an amino acid was present.
Some of the sample and separately authentic glycine was
treated with ninhydrin and K2CO3 in hot water for 1 min. Both
solutions behaved identically and turned a deep red/brown
colour. Both displayed UV spectra with a long wavelength λmax
beyond 500 nm similar to that of an authentic sample of
murexide. The sample was analysed by NMR in an attempt to
quantify a yield for the reaction.An internal standard of sodium
benzoate could be seen showing that there were no paramag-
netic impurities. A sample that had been purified by trituration
with methanol twice showed no signal for glycine at 3.4 ppm.
However crude material after filtration through silica and
drying did display clear evidence for glycine with a peak at
3.4 ppm. The authenticity of this peak was proved by spiking
the NMR tube with glycine. The yield of glycine admixed
with ammonium sulfate was estimated to be 18% for the reac-
tion done at 75 °C and 24% for the room temperature reaction
(Table 1, entries 1 and 2). Some glycine may have been lost in
the work-up since it could bind to the iron salts although treat-
ment of the reaction mixture with excess ammonia prior to
work-up did not increase the quantity of glycine formed
(16% yield, entry 3). However, a reaction performed for 24 h
at room temperature with an excess of ammonia (20 mL) in
water (100 mL) had more glycine present. A yield of 28 and
27% was estimated by H1 NMR (entry 4) from two separate
reactions. The excess of ammonia presumably drives the imine
formation.
However, the previously reported hydrogenation of glyox-
ylic acid with aqNH3 over Pd/C took 6 days before hydrogen
uptake ceased. The slow uptake of hydrogen by a reactive
imine might have been due to its equilibrium with glyoxylic
acid and NH3. We therefore performed another experiment
in which glyoxylic acid was left to stand with Fe(II)SO4 7H2O
and dilute aqNH3 at room temperature for one week. The
presence of glyoxylic acid in the reaction was monitored by
staining TLC plates with Brady’s reagent (eluted with 10%
conc NH3/MeOH). TLC analysis after 24 h of the reactions
heated at 75 °C showed the absence of any glyoxylic acid
which suggested it may have decarboxylated by heating.
After one week at room temperature the yield of glycine in the
product salts was estimated as 20% (entry 5).
Treatment of pyruvic acid with Fe(II)SO4 7H2O and NH3
gave no alanine (analysed by TLC and ninhydrin dip) suggest-
ing that the carbonyl group is not electrophilic enough to react
with ammonia in water at this level of dilution. Pyruvic acid is
a liquid and not a hemi-ketal whereas glyoxylic acid is a solid
with a hydrated aldehyde group.
To investigate the directing effects of a carboxylic acid on
an Fe(II) reduction 2,4-dinitrobenzoic acid was treated with
Fe(II)SO4 7H2O and aqNH3 at 60 °C for 15 min (Scheme 2).
Fe(II)SO4 7H2O and ammonia can reduce nitro groups to
amines,14,15 originally discovered by Gabriel, but this experi-
ment would show if the carboxylic acid group coordinates to
Fe(II) and has a directing effect. Six equivalents of Fe(II)SO4
7H2O were used which is the stoichiometric amount required
to reduce one nitro group to an amine. The product of this
reaction was exclusively 2-amino-4-nitrobenzoic acid17 identi-
fied by its spectroscopic properties, a comparison of them
to authentic commercial material and by its conversion to a
cyclic quinazoline-2,4-dione.
Furthermore a TLC sample of the isomer 4-amino-2-
nitrobenzoic acid was prepared which has a different Rf value.18
The exclusive synthesis of 2-amino-4-nitrobenzoic acid sug-
gests that the carboxylic acid coordinated to the Fe(II) group
and catalysed the reduction of the nitro group. This is also the
more hindered nitro group so coordination to the Fe(II) would
be required for faster reduction. Treatment with NH2NH2 in
EtOH gives the isomer 4-amino-2-nitrobenzoic acid reducing
the less hindered nitro group.18 A second batch of 2-amino-4-
nitrobenzoic acid was obtained by treating the insoluble Fe(III)
salts with aqueous ammonia. This showed that the product
was complexing to the insoluble Fe(III) salts. 2-Amino-4-
nitrobenzoic acid is soluble in methanol (see Experimental)
allowing separation from the iron and ammonium salts.
Glycine from the previous section is harder to purify because
it is poorly soluble in methanol. This experiment provides
precedent that the carboxyl group might facilitate the
reduction of glyoxylic acid imine by complexing to Fe(II). 2-
Amino-4-nitrobenzoic acid was converted to 7-nitroquinazo-
line-2,4-dione by condensation with KOCN following a
preparation for making quinazoline-2,4-dione (Scheme 3).19
6-Nitroquinazoline-2,4-dione has been used in nucleotide
studies and the 7-nitro isomer was previously unknown.20
In summary a potential prebiotic synthesis of the amino acid
glycine is proposed via the reductive amination of glyoxylic
acid with soluble Fe(II)SO4 and ammonia. Arc discharge
Table 1 The maximum theoretical yield is 50% for all entries
using 1 equiv. of Fe(II)SO4 7H2O
Entry Conditions (mL of conc NH3 made up to 100 mL Yield
H2O with 1eqFe(II)SO4 7H2O, temp, time)
/%
1
2
3
1.5 mL conc NH3 75 °C 24 h
18
24
1.5 mL conc NH3 room temperature 24 h
1.5 mL conc NH3 room temperature 24 h work up 18
(20 mL conc NH3)
4
5
20 mL conc NH3 room temperature 24 h
1.5 mL conc NH3 room temperature 7 days
28
20

JOURNAL OF CHEMICAL RESEARCH 2011 131
Glycine (Method 2): Glyoxylic acid hydrate (500 mg, 5.43 mmol)
and Fe(II)SO4 7H2O (1.5 g, 5.39 mmol) were dissolved in water
(100 mL) and treated with concentrated ammonia (1.5 mL). The mix-
ture was left at room temperature with occasional swirling for 24 h
and then worked up as described previously to give an off white solid
(1.17 g). Glycine analysis by 1H NMR (24%).
Glycine (Method 3): Glyoxylic acid hydrate (600 mg, 6.52 mmol)
and Fe(II)SO4 7H2O (1.8 g, 6.48 mmol) were dissolved in water (100
mL) and treated with concentrated ammonia (1.5 mL). The mixture
was heated at 75 °C with occasional swirling for 24 h The mixture
was treated with conc NH3 (20 mL) then worked up as described
previously (1.1 g). Glycine analysis by 1H NMR (18%).
Glycine (Method 4): Glyoxylic acid hydrate (500 mg, 5.43 mmol)
and Fe(II)SO4 7H2O (1.5 g, 5.39 mmol) were dissolved in water
(80 mL) and treated with concentrated ammonia (20 mL). The
mixture was left to stand at room temperature for 24 h and worked
up as described previously to give an off white solid (1.0 g). Glycine
analysis by 1H NMR (28 and 27%).
Glycine (Method 5): Glyoxylic acid hydrate (500 mg, 5.43 mmol)
and Fe(II)SO4 7H2O (1.5 g, 5.39 mmol) were dissolved in water
(100 mL) and treated with concentrated ammonia (1.5 mL). The mix-
ture was stoppered then left to stand at room temperature for 7 days
and worked up as described previously to give an off white solid
(1.1 g). Glycine analysis by 1H NMR (20%).
Scheme 3
experiments through a CO2/N2 atmosphere5 might mimic a
prebiotic flux of amino acids, nitrite and glyoxylic acid. Fe(II)
needed to reduce nitrite to ammonia and for this reduction may
have been present in prebiotic oceans.8 The yield of amino
acids is estimated at only 0.01% in the arc-discharge experi-
ments.5 Although the yields of glycine reported here are higher
the rates will slow down upon dilution. It is thought that to
originate and sustain life a mechanism of pre-concentration of
reagents from a dilute ocean might have occurred.21
Experimental
IR spectra were recorded on an ATI Mattson FTIR spectrometer using
potassium bromide or sodium chloride discs. UV spectra were
recorded using a Perkin-Elmer Lambda 25 UV-VIS spectrometer with
Syntheses
2-Amino-4-nitrobenzoic acid: 2,4-Dinitrobenzoic acid (2.0 g, 9.81
mmol) was added to water (50.0 mL), in a beaker, followed by con-
centrated NH3 (1.0 mL) to aid dissolution. The mixture was heated
to 50 °C to dissolve the solids. Six equivalents of Fe(II)SO4·7H2O
(15.7 g, 56.6 mmol) were added to water (50.0 mL), in a separate
beaker, and heated at 50 °C until dissolved. The two solutions were
combined and concentrated NH3 (20.0 mL) was added slowly, with
stirring. The reaction was left to proceed at 60 °C for 15 min. The
mixture was filtered through a Buchner funnel whilst hot and washed
with water. The red aqueous filtrate was transferred into a round-
bottomed flask, and evaporated in vacuo at 50 °C, leaving the product
as well as the remaining salts. The product was taken up in MeOH
and decanted from the salts into a separate round-bottomed flask, a
procedure, which was repeated until there was no visible sign of the
product in the salts (MeOH stayed clear, and not yellow, when added
to the solid). The brown iron salts, left in the Buchner funnel, were
added to H2O (100 mL) in a beaker and heated to 55 °C. Concentrated
NH3 (20.0 mL) was added to the mixture, with vigorous stirring. The
mixture was left for 10 mins, after which it was filtered and washed
with H2O. The above treatment with concentrated NH3 was repeated
once more on the iron salts. The two resulting aqueous filtrates were
combined and the water was removed in vacuo. The product was taken
up in MeOH and filtered from the remaining iron salts. The methanol
was combined with the first methanol fraction, concentrated in vacuo,
then purified by chromatography on silica gel eluting with MeOH:
DCM 1:9, to give the title compound as a brown solid (1.15 g, 67%)
m.p. 244–245 °C (lit. 264 °C22 and 269 °C17). A separate analysis of
each batch gave firstly 0.75g (44%) and secondly 0.39 g (23%); λmax
(ethanol)/nm 402 (log ε 2.95); νmax (KBr)/cm−1 3510, 3395, 2750,
1680 and 1300; δH (250 MHz; CD3OD) 7.29 (1H, d, J = 8.9 Hz), 7.59
(1H, s) and 7.99 (1H, d, 8.9 Hz); δC (62.9 MHz; CD3OD) 109.6, 111.9,
116.0, 134.4, 152.9, 153.6 and 170.3; m/z (EI) C7H6N2O4 182 (M+
100%); (ASAP) Calcd 183.0400. Found 183.0395 (M+ + H).
7-Nitro-1H-quinazoline-2,4-dione 2-Amino-4-nitrobenzoic acid
(0.25 g, 1.37 mmol) was dissolved in H2O (30.0 mL) and treated with
aqueous NaOH (5.0 mL, 5M). Powdered KOCN (1.0 g, 12.3 mmol)
then glacial acetic acid (2.0 mL) was added to the mixture. The reac-
tion mixture was stirred for 2 h then aqueous HCl (5 mL, 1.0M) was
added. A green precipitate formed which was filtered, dried and
weighed (0.21 g). The filtered solid was dissolved in glacial acetic
acid (75.0 mL), heated to 55 °C and stirred for 3 h. The HOAc was
evaporated on the rotary evaporator and the resulting solid was
dissolved in MeOH. The product was decanted into a separate round-
bottomed flask separating it from impurities. The MeOH was evapo-
rated in vacuo and the mixture purified by chromatography on silica
gel (eluent MeOH 1:9 DCM then MeOH) to give the title compound
(0.16 g, 55%) as a light green/yellow solid, m.p. 178–179 °C; λmax
(ethanol)/nm 328 (log ε 3.2), 353 (3.3) and 383 (3.1); νmax (KBr)/cm−1
3445, 3330, 1685, 1530 and 1305; δH (250 MHz; DMSO-d6) 6.74
EtOH as the solvent. H and 13C NMR spectra were recorded at
1
250 MHz and 62.9 MHz respectively using a Bruker AC 250 spec-
1
trometer. In some cases, H and 13C NMR spectra were recorded at
400 MHz and 100.5 MHz respectively using a Varian 400 spectrome-
ter. Chemical shifts, δ are given in ppm relative to the residual solvent
and coupling constants, J are given in Hz. Low resolution and high
resolution mass spectra were obtained at the University of Wales,
Swansea using electron impact ionisation, chemical ionisation and
electrospray ionisation methods. Melting points were determined on a
Kofler hot-stage microscope.
Glassware was cleaned with conc HCl to remove insoluble Fe(III)
salt residues.
Reductive aminations
Glycine (Method 1): Glyoxylic acid hydrate (600 mg, 6.52 mmol) and
Fe(II)SO4 7H2O (1.8 g, 6.48 mmol) were dissolved in water (100 mL)
and treated with concentrated ammonia (1.5 mL). The mixture was
heated at 75 °C with occasional swirling for 24 h. After cooling and
standing still the mixture was filtered through a 2 cm pad of silica to
give an almost clear solution. The filtration was done by carefully
decanting the solution from the iron salts onto the silica pad because
the iron salts block the filter. The iron salts were then poured onto the
silica pad and washed with a small amount of water. The filtrate gives
a pale yellow solution on concentration in vacuo on a rotary evapora-
tor which was evaporated to dryness at 60–65 °C to give an off-white
solid (1.15 g) consisting of glycine and ammonium sulfate. The lack
of green colouration suggests that no Fe(II) cation is present: λmax
(water)/nm 540–560 (developed by heating some sample with ninhy-
drin and K2CO3 in water in a test tube); νmax (KBr)/cm−1 2800–3600,
1600–1700, 1080–1120 and 600–620 cm−1; δH (400 MHz; D2O) An
expected peak for glycine was found at 3.4 ppm. A sample was spiked
with glycine to prove the authenticity of this peak. The internal stan-
dard of sodium benzoate was observed at 7.28(3H) and 7.68(2H);
Glycine analysis by 1H NMR (18%). TLC on a silica plate showed a
strong glycine spot compared to authentic glycine (2.5 cm plate width,
eluent 10% conc NH3/MeOH, development with a ninhydrin dip of
1 mg mL−1 MeOH). The plate is dried at 100 °C for 1 min after elution
with NH3/MeOH solution to remove traces of ammonia before using
the ninhydrin dip. The ammonia slowly evaporates over a few days
which reduces the Rf value. Initially it is about 0.6–0.8 with the sides
of the spot angled upwards. With less ammonia the Rf value is less and
the sides of the broader spot angle downwards. For some sensitive
reactions using an NH3/MeOH eluent and ninhydrin dip, the wider
silica plate of 2.5 cm is preferred and gives more resolved spots. Note
that traces of salts (such as brine) on TLC spots can decrease the
Rf value of amino acids on silica plates. The glycine in the reaction
mixture sometimes streaks because of the iron salts (depending upon
how much NH3 is left in the eluent) but after filtration through silica
the spot elutes normally.

132 JOURNAL OF CHEMICAL RESEARCH 2011
7
8
9
D.P. Summers and S. Chang, Orig. Life Evolut. Biol., 1994, 24, 99.
H.D. Holland, Econ. Geol., 1973, 68, 1169.
H.D. Holland, The chemical evolution of the atmosphere and oceans,
Princeton University Press, 1984.
(1H, s), 7.70 (1H, d, J = 8.5Hz), 8.13 (1H, d, J = 8.5Hz), 9.30 (1H, s)
and 11.04 (1H, s); δC (62.9 MHz; DMSO-d6) 113.2, 113.8, 122.8,
132.3, 143.4, 149.5, 155.7 and 168.3; m/z (APCI) 208 (M+ 100%);
(APCI) Calcd 208.0353 (M++ H). Found 208.0347.
10 S. Shinkai, H. Hamada, A. Dohyama and O. Manabe, Tetrahedron Lett.,
1980, 21, 1661.
11 M.M. Desnuelle and C. Fromageot, Bull. Soc. Chim. Fr.,(Memoires vol.)
1934, 5, 700.
12 C. Huber and G. Wachtershauser, Tetrahedron Lett., 2003, 44, 1695.
13 F. Tiemann and J. Oppermann, Chem. Ber., 1880, 13, 2057.
14 S. Gabriel, Chem. Ber., 1882, 15, 2291.
Received 3 December 2010; accepted 7 January 2011
Paper 1000460 doi: 10.3184/174751911X12964930076809
Published online: 23 March 2011
15 W.A. Jacobs and M. Heidelberger, J. Am.Chem. Soc., 1917, 39, 1435.
16 A. Eschenmoser, Tetrahedron, 2007, 63, 12821.
17 F.S. Spring and J.C. Woods, J. Chem. Soc., 1945, 625.
18 T. Curtius, J. Fur Prakt. Chem., 1907, 184, 281.
19 N.A. Lange and F.E. Sheibley, Org. Syn. Vol II, 1943, 79.
20 J. Michel, G. Gueguen, J. Vercauteren and S. Moreau, Tetrahedron, 1997,
53, 8457.
References
1
2
3
4
5
A.I. Oparin, Origin of life, Dover publications, 1938.
S.L. Miller, Science, 1953, 117, 528.
J.C.G. Walker, Orig. Life Evolut. Biol., 1985, 16, 117.
J.B.S. Haldane, The rationalist annual, 1929, 3.
H.J. Cleaves, J.H. Chalmers, A. Lazcano, S.L. Miller and J.L. Bada, Orig.
Life Evolut. Biol., 2008, 38, 105.
21 J.P. Ferris and W.J. Hagan, Tetrahedron, 1984, 40, 1093.
22 S.S. Joshi and I.R. Gambhir, J. Am. Chem. Soc., 1956, 78, 2222.
6
D.P. Summers and S. Chang, Nature, 1993, 365, 630.

Copyright of Journal of Chemical Research is the property of Science Reviews 2000 Ltd. and its content may
not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written
permission. However, users may print, download, or email articles for individual use.